]> git.wh0rd.org - tt-rss.git/blobdiff - include/functions.php
remove expandable CDM headlines
[tt-rss.git] / include / functions.php
index 8a869dc08a08aa0318655aea05a08cf077ab5199..1dd9a7a1cc7fd106d5d7a5130bac9612600fe9a6 100755 (executable)
@@ -1,6 +1,6 @@
 <?php
        define('EXPECTED_CONFIG_VERSION', 26);
-       define('SCHEMA_VERSION', 133);
+       define('SCHEMA_VERSION', 134);
 
        define('LABEL_BASE_INDEX', -1024);
        define('PLUGIN_FEED_BASE_INDEX', -128);
@@ -11,8 +11,8 @@
        $fetch_last_error_code = false;
        $fetch_last_content_type = false;
        $fetch_last_error_content = false; // curl only for the time being
+       $fetch_effective_url = false;
        $fetch_curl_used = false;
-       $suppress_debugging = false;
 
        libxml_disable_entity_loader(true);
 
        // default sleep interval between feed updates (sec)
        define_default('MIN_CACHE_FILE_SIZE', 1024);
        // do not cache files smaller than that (bytes)
+       define_default('MAX_CACHE_FILE_SIZE', 64*1024*1024);
+       // do not cache files larger than that (bytes)
+       define_default('MAX_DOWNLOAD_FILE_SIZE', 16*1024*1024);
+       // do not download general files larger than that (bytes)
        define_default('CACHE_MAX_DAYS', 7);
        // max age in days for various automatically cached (temporary) files
        define_default('MAX_CONDITIONAL_INTERVAL', 3600*12);
 
        $schema_version = false;
 
-       function _debug_suppress($suppress) {
-               global $suppress_debugging;
-
-               $suppress_debugging = $suppress;
+       // TODO: compat wrapper, remove at some point
+       function _debug($msg) {
+           Debug::log($msg);
        }
 
-       /**
-        * Print a timestamped debug message.
-        *
-        * @param string $msg The debug message.
-        * @return void
-        */
-       function _debug($msg, $show = true) {
-               global $suppress_debugging;
-
-               //echo "[$suppress_debugging] $msg $show\n";
-
-               if ($suppress_debugging) return false;
-
-               $ts = strftime("%H:%M:%S", time());
-               if (function_exists('posix_getpid')) {
-                       $ts = "$ts/" . posix_getpid();
-               }
-
-               if ($show && !(defined('QUIET') && QUIET)) {
-                       print "[$ts] $msg\n";
-               }
-
-               if (defined('LOGFILE'))  {
-                       $fp = fopen(LOGFILE, 'a+');
-
-                       if ($fp) {
-                               $locked = false;
-
-                               if (function_exists("flock")) {
-                                       $tries = 0;
-
-                                       // try to lock logfile for writing
-                                       while ($tries < 5 && !$locked = flock($fp, LOCK_EX | LOCK_NB)) {
-                                               sleep(1);
-                                               ++$tries;
-                                       }
-
-                                       if (!$locked) {
-                                               fclose($fp);
-                                               return;
-                                       }
-                               }
-
-                               fputs($fp, "[$ts] $msg\n");
-
-                               if (function_exists("flock")) {
-                                       flock($fp, LOCK_UN);
-                               }
-
-                               fclose($fp);
-                       }
-               }
-
-       } // function _debug
-
        /**
         * Purge a feed old posts.
         *
         * @access public
         * @return void
         */
-       function purge_feed($feed_id, $purge_interval, $debug = false) {
+       function purge_feed($feed_id, $purge_interval) {
 
                if (!$purge_interval) $purge_interval = feed_purge_interval($feed_id);
 
 
                CCache::update($feed_id, $owner_uid);
 
-               if ($debug) {
-                       _debug("Purged feed $feed_id ($purge_interval): deleted $rows articles");
-               }
+        Debug::log("Purged feed $feed_id ($purge_interval): deleted $rows articles");
 
                return $rows;
        } // function purge_feed
                }
        }
 
+       // TODO: max_size currently only works for CURL transfers
        // TODO: multiple-argument way is deprecated, first parameter is a hash now
        function fetch_file_contents($options /* previously: 0: $url , 1: $type = false, 2: $login = false, 3: $pass = false,
                                4: $post_query = false, 5: $timeout = false, 6: $timestamp = 0, 7: $useragent = false*/) {
                global $fetch_last_error_content;
                global $fetch_last_content_type;
                global $fetch_last_modified;
+               global $fetch_effective_url;
                global $fetch_curl_used;
 
                $fetch_last_error = false;
                $fetch_last_content_type = "";
                $fetch_curl_used = false;
                $fetch_last_modified = "";
+               $fetch_effective_url = "";
 
                if (!is_array($options)) {
 
                $last_modified = isset($options["last_modified"]) ? $options["last_modified"] : "";
                $useragent = isset($options["useragent"]) ? $options["useragent"] : false;
                $followlocation = isset($options["followlocation"]) ? $options["followlocation"] : true;
+               $max_size = isset($options["max_size"]) ? $options["max_size"] : MAX_DOWNLOAD_FILE_SIZE; // in bytes
+               $http_accept = isset($options["http_accept"]) ? $options["http_accept"] : false;
 
                $url = ltrim($url, ' ');
                $url = str_replace(' ', '%20', $url);
 
                        $ch = curl_init($url);
 
-                       if ($last_modified && !$post_query) {
-                               curl_setopt($ch, CURLOPT_HTTPHEADER,
-                                       array("If-Modified-Since: $last_modified"));
-                       }
+                       $curl_http_headers = [];
+
+                       if ($last_modified && !$post_query)
+                               array_push($curl_http_headers, "If-Modified-Since: $last_modified");
+
+                       if ($http_accept)
+                               array_push($curl_http_headers, "Accept: " . $http_accept);
+
+                       if (count($curl_http_headers) > 0)
+                               curl_setopt($ch, CURLOPT_HTTPHEADER, $curl_http_headers);
 
                        curl_setopt($ch, CURLOPT_CONNECTTIMEOUT, $timeout ? $timeout : FILE_FETCH_CONNECT_TIMEOUT);
                        curl_setopt($ch, CURLOPT_TIMEOUT, $timeout ? $timeout : FILE_FETCH_TIMEOUT);
                        curl_setopt($ch, CURLOPT_ENCODING, "");
                        //curl_setopt($ch, CURLOPT_REFERER, $url);
 
+                       if ($max_size) {
+                               curl_setopt($ch, CURLOPT_NOPROGRESS, false);
+                               curl_setopt($ch, CURLOPT_BUFFERSIZE, 16384); // needed to get 5 arguments in progress function?
+
+                               // holy shit closures in php
+                               // download & upload are *expected* sizes respectively, could be zero
+                               curl_setopt($ch, CURLOPT_PROGRESSFUNCTION, function($curl_handle, $download_size, $downloaded, $upload_size, $uploaded) use( &$max_size) {
+                                       Debug::log("[curl progressfunction] $downloaded $max_size", Debug::$LOG_EXTENDED);
+
+                                       return ($downloaded > $max_size) ? 1 : 0; // if max size is set, abort when exceeding it
+                               });
+
+                       }
+
                        if (!ini_get("open_basedir")) {
                                curl_setopt($ch, CURLOPT_COOKIEJAR, "/dev/null");
                        }
                        $http_code = curl_getinfo($ch, CURLINFO_HTTP_CODE);
                        $fetch_last_content_type = curl_getinfo($ch, CURLINFO_CONTENT_TYPE);
 
+                       $fetch_effective_url = curl_getinfo($ch, CURLINFO_EFFECTIVE_URL);
+
                        $fetch_last_error_code = $http_code;
 
                        if ($http_code != 200 || $type && strpos($fetch_last_content_type, "$type") === false) {
 
                         $context_options = array(
                                  'http' => array(
+                                               'header' => array(
+                                                       'Connection: close'
+                                               ),
                                                'method' => 'GET',
                                                'ignore_errors' => true,
                                                'timeout' => $timeout ? $timeout : FILE_FETCH_TIMEOUT,
                                                'protocol_version'=> 1.1)
                                  );
 
-                       if (!$post_query && $last_modified) {
-                               $context_options['http']['header'] = "If-Modified-Since: $last_modified\r\n";
-                       }
+                       if (!$post_query && $last_modified)
+                               array_push($context_options['http']['header'], "If-Modified-Since: $last_modified");
+
+                       if ($http_accept)
+                               array_push($context_options['http']['header'], "Accept: $http_accept");
 
                        if (defined('_HTTP_PROXY')) {
                                $context_options['http']['request_fulluri'] = true;
 
                        $old_error = error_get_last();
 
+                       $fetch_effective_url = $url;
+
                        $data = @file_get_contents($url, false, $context);
 
                        if (isset($http_response_header) && is_array($http_response_header)) {
                                                        // e.g. if we were being redirected -- last one is the right one
                                                } else if ($key == 'last-modified') {
                                                        $fetch_last_modified = $value;
+                                               } else if ($key == 'location') {
+                                                       $fetch_effective_url = $value;
                                                }
                                        }
 
                $profile = $profile ? $profile : null;
 
                $u_sth = $pdo->prepare("SELECT pref_name
-                       FROM ttrss_user_prefs WHERE owner_uid = :uid AND 
+                       FROM ttrss_user_prefs WHERE owner_uid = :uid AND
                                (profile = :profile OR (:profile IS NULL AND profile IS NULL))");
                $u_sth->execute([':uid' => $uid, ':profile' => $profile]);
 
 
                if (!SINGLE_USER_MODE) {
                        $user_id = false;
+                       $auth_module = false;
 
                        foreach (PluginHost::getInstance()->get_hooks(PluginHost::HOOK_AUTH_USER) as $plugin) {
 
                                $user_id = (int) $plugin->authenticate($login, $password);
 
                                if ($user_id) {
-                                       $_SESSION["auth_module"] = strtolower(get_class($plugin));
+                                       $auth_module = strtolower(get_class($plugin));
                                        break;
                                }
                        }
 
                        if ($user_id && !$check_only) {
-                               @session_start();
+
+                               session_start();
+                               session_regenerate_id(true);
 
                                $_SESSION["uid"] = $user_id;
                                $_SESSION["version"] = VERSION_STATIC;
+                               $_SESSION["auth_module"] = $auth_module;
 
                                $pdo = DB::pdo();
                                $sth = $pdo->prepare("SELECT login,access_level,pwd_hash FROM ttrss_users
        }
 
        function logout_user() {
-               session_destroy();
+               @session_destroy();
                if (isset($_COOKIE[session_name()])) {
                   setcookie(session_name(), '', time()-42000, '/');
                }
+               session_commit();
        }
 
        function validate_csrf($csrf_token) {
                                }
 
                                if (!$_SESSION["uid"]) {
-                                       @session_destroy();
-                                       setcookie(session_name(), '', time()-42000, '/');
+                                       logout_user();
 
                                        render_login_form();
                                        exit;
 
                                /* cleanup ccache */
 
-                               $sth = $pdo->prepare("DELETE FROM ttrss_counters_cache WHERE owner_uid = ? 
+                               $sth = $pdo->prepare("DELETE FROM ttrss_counters_cache WHERE owner_uid = ?
                                        AND
                                                (SELECT COUNT(id) FROM ttrss_feeds WHERE
                                                        ttrss_feeds.id = feed_id) = 0");
 
                                $sth->execute([$_SESSION['uid']]);
 
-                               $sth = $pdo->prepare("DELETE FROM ttrss_cat_counters_cache WHERE owner_uid = ? 
+                               $sth = $pdo->prepare("DELETE FROM ttrss_cat_counters_cache WHERE owner_uid = ?
                                        AND
                                                (SELECT COUNT(id) FROM ttrss_feed_categories WHERE
                                                        ttrss_feed_categories.id = feed_id) = 0");
                                "feed_debug_viewfeed" => __("Debug viewfeed()"),
                                "catchup_all" => __("Mark all feeds as read"),
                                "cat_toggle_collapse" => __("Un/collapse current category"),
-                               "toggle_combined_mode" => __("Toggle combined mode"),
-                               "toggle_cdm_expanded" => __("Toggle auto expand in combined mode")),
+                               "toggle_combined_mode" => __("Toggle combined mode")),
                        __("Go to") => array(
                                "goto_all" => __("All articles"),
                                "goto_fresh" => __("Fresh"),
                        "f *d" => "feed_debug_update",
                        "f *g" => "feed_debug_viewfeed",
                        "f *c" => "toggle_combined_mode",
-                       "f c" => "toggle_cdm_expanded",
                        "*q" => "catchup_all",
                        "x" => "cat_toggle_collapse",
        //                      "goto" => array(
                $data["num_feeds"] = (int) $num_feeds;
 
                $data['last_article_id'] = Article::getLastArticleId();
-               $data['cdm_expanded'] = get_pref('CDM_EXPANDED');
 
                $data['dep_ts'] = calculate_dep_timestamp();
                $data['reload_on_ts_change'] = !defined('_NO_RELOAD_ON_TS_CHANGE');
                $search_query_leftover = array();
 
                $pdo = Db::pdo();
-               
+
                if ($search_language)
                        $search_language = $pdo->quote(mb_strtolower($search_language));
                else
                return false;
        }
 
-       function sanitize($str, $force_remove_images = false, $owner = false, $site_url = false, $highlight_words = false, $article_id = false) {
-               if (!$owner) $owner = $_SESSION["uid"];
-
-               $res = trim($str); if (!$res) return '';
+       // check for locally cached (media) URLs and rewrite to local versions
+       // this is called separately after sanitize() and plugin render article hooks to allow
+       // plugins work on original source URLs used before caching
 
+       function rewrite_cached_urls($str) {
                $charset_hack = '<head>
                                <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
                        </head>';
 
-               $res = trim($res); if (!$res) return '';
-
-               libxml_use_internal_errors(true);
+               $res = trim($str); if (!$res) return '';
 
                $doc = new DOMDocument();
                $doc->loadHTML($charset_hack . $res);
                $xpath = new DOMXPath($doc);
 
-               $rewrite_base_url = $site_url ? $site_url : get_self_url_prefix();
+               $entries = $xpath->query('(//img[@src]|//video[@poster]|//video/source[@src]|//audio/source[@src])');
 
-               $entries = $xpath->query('(//a[@href]|//img[@src]|//video/source[@src]|//audio/source[@src])');
+               $need_saving = false;
 
                foreach ($entries as $entry) {
 
-                       if ($entry->hasAttribute('href')) {
-                               $entry->setAttribute('href',
-                                       rewrite_relative_url($rewrite_base_url, $entry->getAttribute('href')));
+                       if ($entry->hasAttribute('src') || $entry->hasAttribute('poster')) {
 
-                               $entry->setAttribute('rel', 'noopener noreferrer');
-                       }
-
-                       if ($entry->hasAttribute('src')) {
-                               $src = rewrite_relative_url($rewrite_base_url, $entry->getAttribute('src'));
+                               // should be already absolutized because this is called after sanitize()
+                               $src = $entry->hasAttribute('poster') ? $entry->getAttribute('poster') : $entry->getAttribute('src');
                                $cached_filename = CACHE_DIR . '/images/' . sha1($src);
 
                                if (file_exists($cached_filename)) {
 
                                        $src = get_self_url_prefix() . '/public.php?op=cached_url&hash=' . sha1($src) . $suffix;
 
-                                       if ($entry->hasAttribute('srcset')) {
-                                               $entry->removeAttribute('srcset');
-                                       }
+                                       if ($entry->hasAttribute('poster'))
+                                               $entry->setAttribute('poster', $src);
+                                       else
+                                               $entry->setAttribute('src', $src);
 
-                                       if ($entry->hasAttribute('sizes')) {
-                                               $entry->removeAttribute('sizes');
-                                       }
+                                       $need_saving = true;
                                }
+                       }
+               }
+
+               if ($need_saving) {
+                       $doc->removeChild($doc->firstChild); //remove doctype
+                       $res = $doc->saveHTML();
+               }
+
+               return $res;
+       }
+
+       function sanitize($str, $force_remove_images = false, $owner = false, $site_url = false, $highlight_words = false, $article_id = false) {
+               if (!$owner) $owner = $_SESSION["uid"];
+
+               $res = trim($str); if (!$res) return '';
+
+               $charset_hack = '<head>
+                               <meta http-equiv="Content-Type" content="text/html; charset=utf-8"/>
+                       </head>';
+
+               $res = trim($res); if (!$res) return '';
+
+               libxml_use_internal_errors(true);
+
+               $doc = new DOMDocument();
+               $doc->loadHTML($charset_hack . $res);
+               $xpath = new DOMXPath($doc);
+
+               $rewrite_base_url = $site_url ? $site_url : get_self_url_prefix();
+
+               $entries = $xpath->query('(//a[@href]|//img[@src]|//video/source[@src]|//audio/source[@src])');
+
+               foreach ($entries as $entry) {
+
+                       if ($entry->hasAttribute('href')) {
+                               $entry->setAttribute('href',
+                                       rewrite_relative_url($rewrite_base_url, $entry->getAttribute('href')));
+
+                               $entry->setAttribute('rel', 'noopener noreferrer');
+                       }
+
+                       if ($entry->hasAttribute('src')) {
+                               $src = rewrite_relative_url($rewrite_base_url, $entry->getAttribute('src'));
+
+                               // cache stuff has gone to rewrite_cached_urls()
 
                                $entry->setAttribute('src', $src);
                        }
                        if ($entry->nodeName == 'img') {
                                $entry->setAttribute('referrerpolicy', 'no-referrer');
 
+                               $entry->removeAttribute('width');
+                               $entry->removeAttribute('height');
+
                                if ($entry->hasAttribute('src')) {
                                        $is_https_url = parse_url($entry->getAttribute('src'), PHP_URL_SCHEME) === 'https';
 
                                                }
                                        }
                                }
+                       }
+
+                       if ($entry->hasAttribute('src') &&
+                                       ($owner && get_pref("STRIP_IMAGES", $owner)) || $force_remove_images || $_SESSION["bw_limit"]) {
+
+                               $p = $doc->createElement('p');
+
+                               $a = $doc->createElement('a');
+                               $a->setAttribute('href', $entry->getAttribute('src'));
+
+                               $a->appendChild(new DOMText($entry->getAttribute('src')));
+                               $a->setAttribute('target', '_blank');
+                               $a->setAttribute('rel', 'noopener noreferrer');
 
-                               if (($owner && get_pref("STRIP_IMAGES", $owner)) ||
-                                       $force_remove_images || $_SESSION["bw_limit"]) {
+                               $p->appendChild($a);
 
-                                       $p = $doc->createElement('p');
+                               if ($entry->nodeName == 'source') {
 
-                                       $a = $doc->createElement('a');
-                                       $a->setAttribute('href', $entry->getAttribute('src'));
+                                       if ($entry->parentNode && $entry->parentNode->parentNode)
+                                               $entry->parentNode->parentNode->replaceChild($p, $entry->parentNode);
 
-                                       $a->appendChild(new DOMText($entry->getAttribute('src')));
-                                       $a->setAttribute('target', '_blank');
-                                       $a->setAttribute('rel', 'noopener noreferrer');
+                               } else if ($entry->nodeName == 'img') {
 
-                                       $p->appendChild($a);
+                                       if ($entry->parentNode)
+                                               $entry->parentNode->replaceChild($p, $entry);
 
-                                       $entry->parentNode->replaceChild($p, $entry);
                                }
                        }
 
                        }
                }
 
-               $allowed_elements = array('a', 'address', 'acronym', 'audio', 'article', 'aside',
+               $allowed_elements = array('a', 'abbr', 'address', 'acronym', 'audio', 'article', 'aside',
                        'b', 'bdi', 'bdo', 'big', 'blockquote', 'body', 'br',
                        'caption', 'cite', 'center', 'code', 'col', 'colgroup',
                        'data', 'dd', 'del', 'details', 'description', 'dfn', 'div', 'dl', 'font',
        }
 
        function tag_is_valid($tag) {
-               if ($tag == '') return false;
-               if (is_numeric($tag)) return false;
-               if (mb_strlen($tag) > 250) return false;
-
-               if (!$tag) return false;
+               if (!$tag || is_numeric($tag) || mb_strlen($tag) > 250)
+                       return false;
 
                return true;
        }
                }
 
                $sth = $pdo->prepare("SELECT id FROM ttrss_feed_categories
-                               WHERE (parent_cat = :parent OR (:parent IS NULL AND parent_cat IS NULL)) 
+                               WHERE (parent_cat = :parent OR (:parent IS NULL AND parent_cat IS NULL))
                                AND title = :title AND owner_uid = :uid");
                $sth->execute([':parent' => $parent_cat_id, ':title' => $feed_cat, ':uid' => $_SESSION['uid']]);
 
                $sth = $pdo->prepare("SELECT access_key FROM ttrss_access_keys
                                WHERE feed_id = ? AND is_cat = ?
                                AND owner_uid = ?");
-               $sth->execute([$feed_id, (int)$is_cat, $owner_uid]);
+               $sth->execute([$feed_id, $is_cat, $owner_uid]);
 
                if ($row = $sth->fetch()) {
                        return $row["access_key"];
                                        (access_key, feed_id, is_cat, owner_uid)
                                        VALUES (?, ?, ?, ?)");
 
-                       $sth->execute([$key, $feed_id, (int)$is_cat, $owner_uid]);
+                       $sth->execute([$key, $feed_id, $is_cat, $owner_uid]);
 
                        return $key;
                }
        }
 
        function get_minified_js($files) {
-               require_once 'lib/jshrink/Minifier.php';
 
                $rv = '';
 
                should be loaded systemwide in config.php */
        function send_local_file($filename) {
                if (file_exists($filename)) {
+
+                       if (is_writable($filename)) touch($filename);
+
                        $tmppluginhost = new PluginHost();
 
                        $tmppluginhost->load(PLUGINS, PluginHost::KIND_SYSTEM);
                        }
 
                        $mimetype = mime_content_type($filename);
+
+                       // this is hardly ideal but 1) only media is cached in images/ and 2) seemingly only mp4
+                       // video files are detected as octet-stream by mime_content_type()
+
+                       if ($mimetype == "application/octet-stream")
+                               $mimetype = "video/mp4";
+
                        header("Content-type: $mimetype");
 
                        $stamp = gmdate("D, d M Y H:i:s", filemtime($filename)) . " GMT";